A Brand Owner’s View
Senior Director of Global Media, Adidas
Over the past few years, there’s been a rallying cry for brands to rebalance their focus and address both the long and short term.
Despite the seminal work from Binet and Field and the attempts by the IPA, we’re trapped in a performance cycle where we can’t look beyond optimising the present – with demands that feel like they are progressively smothering us. Clearly, this is not the optimum environment for creativity to thrive in. Unless we want everything to be relegated to a six-second pre-roll, we need to radically re-think our approach.
I’m not sure the battle is ‘brand’ vs. ‘performance’ – although friction between the two sets of teams involved is often prevalent. Rather, the problem derives from a conflation of misunderstandings and demands, sometimes flowing from the top.
An obsession with accountability
Digital is not a media channel, it is an infrastructure that most media channels utilise. What we really mean when we say ‘digital media’ is desktop and mobile. As these are both inherently accountable, there is a wealth of data on how they perform.
If tags from an ad server are placed on a website and that website has an eCommerce facility, there is also a direct numerical link from the performance of a desktop or mobile advert to sales. How we apportion this is known as attribution, and one of the most widely-used models of the six (or more) in use is ‘last non-direct click’.
In one situation I’ve seen, using two different attribution models from the same system, credit fluctuated by €150 million in sales across a 12-month period.
Other problems with attribution include branded search terms taking disproportionate credit and the self-serving bias of only crediting desktop and mobile with driving online sales. The last point begs the question: 'Why does Amazon spend so much money on TV advertising?'
When comparing attribution to other methods like market mix modelling, its reliability becomes even more questionable. MMM tends to look more holistically over longer periods than attribution.
It covers a wider range of marketing tactics as well as all media channels – meaning it is likely to be more representative. Often, the analysis coming from MMM directly contradicts the analysis coming from attribution. So, which is true?
The answer appears to be neither. Both models are flawed, as both are incomplete. Whilst MMM is more holistic, it suffers from being built on correlation to forecast causation.
Arguably, the right thing to do is to take the results from both systems and test them over both the long and short term, rather than what appears to be more common: choose one favoured system and call it the truth.
Short-termism by design
One thing MMM does not have in its favour when compared to attribution is time lag. Attribution windows are short, often real-time, as they assume people are in market to buy products now. By contrast, MMM takes longer to receive results, in part because it assumes people are not in market to buy when a brand advertises.
The latter is closer to the idea of what we call 'mental availability and priming'.
MMM time lags aren’t conducive to quarterly investor calls. And they do not work seamlessly with organisations that structure teams around agile, sprints and OKRs. Attribution, digital analytics and dashboards, however, all do. As digital transformation projects come to an end, the post-digital age looks very much like a short-termist one.
Even within tactics, there is favouritism and bias, and anecdotally, I have seen this bias manifest. A while back, an internal team obfuscated results from retargeting tests, where incrementality was being assessed.
The team produced data which showed a certain technology had driven 100% incrementality. When further analysed by non-team members, the data showed incrementality was less than 2%.
What is clear to me is that most brands are too heavily focused on short-term over long-term growth, especially if this is driven from the top. Our industry will be much less interesting as a result – and consumers will switch off.
It also appears to me that as organisations have looked to modernise, teams have inadvertently been set up to compete against each other, rather than integrate towards a common goal in an agnostic and balanced approach.
And if we don’t redress the imbalance, there is likely to be no short term to draw from. Brands could be eroded to commodities. We need to run both long-term brand-building and short-term activation.
Like most things, it’s about getting the balance right. Let’s swing the pendulum back towards a longer-term mindset where brands can flourish.